The Dynamic Field Theory (DFT) is a neural network model of spatial workung memory (SWM) first proposed by Thelen et al. (2001) to explain the disappearance of the A-not-B error in infancy. The goal of the proposed research is to test a new version of the DFT (Spencer et al., 2002) that explains the processes underlying a qualitative change in SWM in childhood. Previous research has found that between three and six years of age, there is a qualitative change in how children use reference axes when remembering locations in space. This research will examine whether the change in reference axis effects can be explained with a quantitative change in the same parameters Thelen et al. used to explain the disappearance of the A-not-B error; thereby, connecting a transition in infancy with a transition later in development. Experiment 1 explores the timing of the transition in reference axis effects and the nature of the transition. Experiment 2 tests thr prediction of the DFT that increasing the salience of a reference axis will cause children who are close to the transition to use the reference axis in the same way as older children. Lastly, Experiment 3 uses microgenetic methods to examine the role of experience in this transition. When completed, this research will have moved the DFT closer to being a complete account of the development of spatial working memory. Moreover, this research will have taken an important step forward in the study of the mechanisms underlying developmental change.